tcp: TCP Small Queues

This introduce TSQ (TCP Small Queues)

TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.

sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.

TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.

As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.

This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.

Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.

Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)

I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.

As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.

If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.

[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
  but some drivers call it in their start_xmit() handler.
  These drivers should at least use BQL, or else a single TCP
  session can still fill the whole NIC TX ring, since TSQ will
  have no effect.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 14 insertions, 0 deletions

diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index 47b6c79e9b05..e20c17a7d34e 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -551,6 +551,20 @@ tcp_thin_dupack - BOOLEAN
         Documentation/networking/tcp-thin.txt
         Default: 0
 
+tcp_limit_output_bytes - INTEGER
+        Controls TCP Small Queue limit per tcp socket.
+        TCP bulk sender tends to increase packets in flight until it
+        gets losses notifications. With SNDBUF autotuning, this can
+        result in a large amount of packets queued in qdisc/device
+        on the local machine, hurting latency of other flows, for
+        typical pfifo_fast qdiscs.
+        tcp_limit_output_bytes limits the number of bytes on qdisc
+        or device to reduce artificial RTT/cwnd and reduce bufferbloat.
+        Note: For GSO/TSO enabled flows, we try to have at least two
+        packets in flight. Reducing tcp_limit_output_bytes might also
+        reduce the size of individual GSO packet (64KB being the max)
+        Default: 131072
+
 UDP variables:
 
 udp_mem - vector of 3 INTEGERs: min, pressure, max